I am struggling to figure out the most appropriate way to analyze the data I have collected. I will provide a brief overview. This is for my dissertation.
I have 70+ subjects that I collected data on pre-intervention, I then applied the intervention and recollected data. The problem I am having is the data is not numerical, I can't give specifics but this would be an example of similar info:
All subjects told me their two favorite colors, I then showed them a video describing how purple and red are the best colors, after the video I questioned each subject a second time to see if they keep their original favorite colors or if they changed to red and purple.
How do I assess for change in choice as a whole group? I originally went with a McNemar's test but I am not sure that is correct.
You are making two measurements on the same sample, so I think the best method is agreement analysis, using kappa coefficient.
You should make a 2 x 2 table (columns: initial response; rows: second response), and use kappa coefficient.
This is different than concordance and I think that more appropriate for your design.
Here is a reference: Method agreement analysis: A review of correct methodology
P.F.WatsonaA.Petrieb.https://doi.org/10.1016/j.theriogenology.2010.01.003
Best regards.
Julio.
You are making two measurements on the same sample, so I think the best method is agreement analysis, using kappa coefficient.
You should make a 2 x 2 table (columns: initial response; rows: second response), and use kappa coefficient.
This is different than concordance and I think that more appropriate for your design.
Here is a reference: Method agreement analysis: A review of correct methodology
P.F.WatsonaA.Petrieb.https://doi.org/10.1016/j.theriogenology.2010.01.003
Not sure what happened with Julio and Nasser's responses? Nasser's seems to be the same as Julio's.
Tanya - why can't you tell us your study specifics?
Anyway - you could make life easy for yourself and just use simple, descriptive measures i.e. Likert and/or semantic differential - and then describe variance over the period.
I think he may have copied and pasted my answer by mistake.
Julio.
Dean
Hello and thanks for the response. My study uses a proprietary tool so it makes it difficult to explain.
Before starting research, there needs to a be problem statement (what problems are required to be solved) and the analysis should be done as to provide solutions accordingly. This also indicates the potential applications of the research.
- Badges
- Science topic
- Similar topics
- Medical Science Nursing
- Dissertations